Buffer Composition Recipe Determination via Predictive pH and Conductivity Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining buffer composition recipes for chromatography are inefficient, requiring substantial resources and compromising accuracy when dealing with salts, additives, and multiple buffers at different concentrations, leading to wasted materials and suboptimal results.
Innovation Solution
A method involving design of experiment (DoE) data to select unique buffer compositions, running experiments to obtain results, and using prediction data to optimize pH and conductivity values, thereby determining optimal buffer composition recipes that reduce waste and improve chromatography outcomes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If buffer composition recipes are determined through testing and experimentation, then the desired pH and conductivity values can be achieved, but substantial resources are wasted in the form of man-hours and buffer components
Solution Approach 1:
The system performs preliminary calculations using a predictive model to determine the optimal buffer composition recipe before actual mixing occurs. The model predicts pH and conductivity values based on buffer component concentrations, allowing the system to identify the optimal recipe in advance and avoid wasteful trial-and-error experimentation
Solution Approach 2:
The system incorporates feedback mechanisms where actual pH and conductivity measurements from previous experiments are used to refine and retrain the predictive model. This continuous feedback loop improves prediction accuracy over time, reducing the need for extensive testing and minimizing waste of buffer components
2Extent of automation
If commercial software systems are used to determine buffer composition recipes, then recipe determination is automated, but accuracy is compromised when salts, additives and multiple buffers are used at different concentrations
Solution Approach 1:
The system dynamically adjusts prediction parameters based on the specific buffer composition being analyzed. When salts, additives, or multiple buffers are detected in the formulation, the model automatically modifies its prediction parameters and calculation methods to account for these complex interactions, maintaining high accuracy across diverse buffer types
Solution Approach 2:
The system continuously self-improves by automatically training its predictive model on new experimental data. The model adapts to specific laboratory conditions, buffer components, and interactions autonomously, enhancing accuracy for complex formulations involving salts and additives without requiring manual reconfiguration
3Measurement precision
If extensive testing is performed to determine buffer composition recipes, then accurate pH and conductivity values can be achieved, but the time and complexity of the process increases substantially
Solution Approach 1:
The system performs preliminary predictions using the trained model to identify the most promising buffer recipes before conducting actual experiments. This preliminary screening reduces the number of full-scale tests needed, achieving accurate results faster by focusing experimental efforts only on the most likely optimal compositions
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention relates to an apparatus, a method, a computer program and a computer program product for determining one or more buffer composition recipes. The method comprising obtaining design of experiment, DoE, data, wherein the DoE data is indicative of a set of buffer compositions and corresponding unique recipes, wherein the set of buffer compositions is selected as a subset from a total set of buffer compositions within a design range, running a first set of experiments by consecutively providing buffer compositions mixed according to each unique recipe indicated by the DoE data as the only input to the chromatography apparatus, obtaining results of experiment, RoE, data as output from the chromatography apparatus, wherein the RoE data is indicative of at least a potential of hydrogen, p H, value and a conductivity value of each buffer composition of the set of buffer compositions, obtaining prediction of experimental, PoE data indicative of at least a predicted pH value and a predicted conductivity value of each buffer composition of the total set of buffer compositions, obtaining a first objective function, the first objective function being dependent on a pH value and a conductivity value, selecting a second subset from the total set of buffer compositions which corresponding p H values and conductivity values optimize the first objective function, determining the one or more buffer composition recipes for chromatography of a chemical sample as the unique recipes corresponding to the second subset. The present invention further relates to an apparatus, a computer program and a computer program product.